Ten Years of Permeable Reactive Barriers ... - ACS Publications

Chapter 3

Ten Years of Permeable Reactive Barriers: Lessons Learned and Future Expectations 1

Downloaded by PURDUE UNIV on July 6, 2016 | http://pubs.acs.org Publication Date: November 10, 2002 | doi: 10.1021/bk-2002-0837.ch003

Scott D. Warner and Dominique Sorel 3

2

th

Geomatrix Consultants, 2101 Webster Street, 12 Floor, Oakland, CA 94612 (telephone: 510-663-4100; [email protected]) S.S. Papadopulos & Associates, 217 Church Street, San Francisco, CA 94111 2

The permeable reactive barrier (PRB) has been an innovative in situ ground water remediation concept for more than 15 years, though it has been only over the last 10 years that the PRB has been considered a practical alternative for groundwater treatment. Although close to 50 full-scale and tens more pilot tests have been installed since the first commercial application in 1994, and there have been close to one thousand technical papers and presentations since the early 1990s, the technology is still not considered a developed technology due to a perceived lack of cost and performance data. Over the past ten years, the number of full-scale PRBs installed may have been limited first by the uncer taintyin applying a new remedial technology, and second, by the need for the designer and user to collect and understand the comprensive characterization, engineering, and logistical information required for each site specific considera tion.However, the use of the PRB will become more commonplace as alterna tive materials and installation options continue to develop, and more information on successes, failures, and cost information become available.

36

© 2003 American Chemical Society Henry and Warner; Chlorinated Solvent and DNAPL Remediation ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

37


INTRODUCTION The permeable reactive barrier (PRB) has been a lexicon of the groundwater remediation industry for nearly a decade. Described as a concept more than 15 years ago by McMurty and Elton (1), this relatively simple concept of passively treating groundwater contaminants using a subsurface zone of reactive material was first subject to practical field demonstrations at the Canadian Base Borden site in 1991 by researchers from the University of Waterloo (2, 3, 4). The first commercial application of the PRB as a final groundwater remedy came three years later with the installation of the PRB at a former manufacturing site in Sunnyvale, California (5, 6). Since that time, PRBs have been installed at close to 50 sites as full-scale groundwater remedies to treat a variety of organic and inorganic chemical constituents. Also, nearly 20 additional pilot tests of PRBs have been completed or are ongoing. Yet, the PRB concept continues to be characterized as innovative (rather than developed) due to a perceived lack of data on cost and performance and a consideration that its application remains limited (7). Certainly, the PRB concept first gained acceptance due to the perception that passively treating groundwater in situ without the use of pumps or other energy-intensive devises, and without the long-term operations and maintenance demands associated with active treatment methods such as "pump-and-treat," would result in a favorable cost to benefit remedy. For some sites, including the Sunnyvale application, this remains true; for other sites, the applications have either been too recent to accurately assess cost and performance information, or uncertainties in performance may exist. The purpose of this paper is to recount and assess some aspects of the evolution of the PRB as a remediation concept as it moves into the next decade of development and application. In doing so, we attempt to focus on lessons learned and thus lay the groundwater for assuring die proper and successful use of this remedial concept.

CONCEPT AND RESOURCES The concept of destroying or immobilizing groundwater contaminants using a subsurface treatment zone, as described by McMurty and Elton (1), is a relatively simple concept by which a material capable of directly (or indirectly through geochemical and biological enhancement of the ambient system) destroying or immobilizing the target chemical constituents is placed in the sub-

Henry and Warner; Chlorinated Solvent and DNAPL Remediation ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

38


surface to intersect the contaminant plume flowing typically under its own hydraulic gradient. As shown in Figure 1, the PRB could be oriented to capture and mitigate a plume migrating both laterally, as well as vertically downward, though the practical application has been to affect laterally-spreading plumes. The PRB can be placed immediately downgradient of a chemical plume to prevent the plume from migrating further, or immediately downgradient of the contaminant source to prevent a plume from developing. The typical PRB is designed not to impede groundwater flow, although ambient hydraulic conditions can be altered by the PRB system, and such potential hydraulic effects must be considered in the design to avoid unintended performance.

Figure 1. Typical two-dimensional representation of a permeable reactiv barrier or applied reactive treatment zone (ARTZ); (a) vertical orientatio lateral capture; (b) horizontal orientation for capture of downward migra chemicals. We consider the PRB to be a general category of remediation solutions rather than a specific treatment device because each PRB must be designed for site specific geological, hydraulic, chemical, structural, legal, and economic (e.g., land use) conditions (8). Consequently, consideration of the geometry, design, treatment media, and construction method of each PRB is unique. The consideration of the many parameters necessary to assure a successful deployment of a PRB somewhat explains the difficulty that potential users and regulators have in determining whether a PRB is appropriate for a given site, and in evaluating design criteria prior to implementation.


39 The site specific uniqueness and consideration of many PRB design variables also partially explains why, since 1991, several hundred technical papers, abstracts, conference proceedings, and more recently, internet web sites, have reported on the development and research of PRBs including field applications, and the chemical and biological reactions of various proposed treatment media. One of the first major conferences to formally devote a session to PRB and related studies was the April 1995 American Chemical Society (ACS) 209 National Meeting held in Anaheim, California. Approximately 40 papers were presented with the focus generally concerning research on the use of the most popular reactive treatment media, zero-valent metals and chiefly zero-valent granular iron, to promote contaminant remediation. Though conferences devoted to in situ remediation methods had been held in previous years (9), the 1995 ACS meeting appeared to catalyze industry attention on PRBs as a serious remedial alternative. Since then there have been no less than ten major technical conferences with formal sessions on PRBs and related treatment media (including the 1997 and 2001 ACS national meetings, the 1998 and 2000 Battelle conferences on remediation of chlorinated and recalcitrant compounds, and the 1995, 1997, and 2001 International Containment Technology Conferences).


th

Table 1 provides a selected list of internet resources available on PRB related topics. These resources are useful for understanding the development of PRBs and treatment media, and are becoming more important for potential users, regulators, and the general public in assessing appropriate PRB design and application. Of note include two internet web sites - the iron treatment media references data base maintained by the Oregon Graduate Institute (OGI), and the Remediation Technology Development Forum (RTDF) - which are devoted to providing technical references and case study information on the use of zero-valent metals as a treatment media, and on PRB deployment, respectively. As of the start of 2001, nearly 500 technical references are listed on the OGI data base and more than 160 references are provided on the RTDF site. The growth in PRB related publication activity is demonstrated in part by the history of publication dates listed on the OGI data base (Figure 3). PHYSICAL AND C H E M I C A L TREATMENT CONSIDERATIONS The PRB contains two principal components: the treatment matrix, and the hydraulic control system. The treatment matrix is the material or the zone where treatment takes place. Table 2 provides a partial list of treatment materials that have been used or considered for PRB use. The identification and development of alternative treatment materials is becoming a regular occurrence to better


40

Table 1. Selected Internet Sites on PRBs and Treatment Media Web Site Name

Web Site Address


Oregon Graduate Institute http://cgr.ese.ogi.edu/iron/ Remediation with Iron Searchable database on remediation by zero-valent metals Remediation Technology http://www.rtdf.org/ Development Forum Permeable Barriers Subgroup Technical literature and case studies on deployment of permeable reactive barriers Groundwater Remediation http://www.gwrtac.org/ Technology Analysis Center Technical literature including technical status reports and case studies Interstate Technology Regulatory http://www.itrcweb.org/ Cooperation Guidance documents for PRB design prepared by a coalition of state regulatory agencies U.S. EPA Technology Innovation http://www.clu-in.org/ Office Technical literature including status reports, case studies, reference lists

remedy complex contaminant mixtures, as well as enhancing the cost effectiveness and longevity of treatment options. For example, Taylor, et al. (10) described laboratory tests evaluating a variety of potential treatment media, including polymer-coated-basalt, enhanced apatite, and crushed pecan shells to mitigate radionuclides. The treatment zone itself may be a trench excavated and subsequently filled with a reactive material, a zone of treatment material that has emplaced via injection, or a zone of geochemically altered native material formed via flushing or injecting a chemical substrate (liquid, solid, or gaseous). Most PRBs have been implemented to treat groundwater affected by chlorinated hydrocarbon compounds (13). The reactive material of choice for this treatment has been granular zero-valent iron (Figure 2). This treatment material has been the focus of most technical papers on PRBs as demonstrated by comparing paper titles from the 1995 ACS session with this 2001 ACS session which shows that more than 40 of the 50 presentations from the combined sessions are focused on the use of iron as the primary treatment media. What is different today versus 5 to 10 years ago, however, is more a function of practical and economic needs (e.g, lower material costs, assessing methods to increase longevity of reactive materials, developing less intrusive implementation methods, developing deeper implementation methods, etc.) although, there have been some significant developments with respect to understanding the reaction Henry and Warner; Chlorinated Solvent and DNAPL Remediation ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

41


Table 2. Partial List of Reactive Materials used in PRBs

Sources: USEPA (11), USDOE-Grand Junction (12).



42

Figure 2. Examples of treatment media used in PRB applications: (I) granular iron metal (left), and (2) the zeolite clinoptilolite (right). mechanisms involved with iron (as well as other reactive materials). For example, although the treatment of chlorinated hydrocarbons was initially thought to rely on the sequential dehalogenation of higher chlorinated compounds (3), the basic premise today is that destruction of the target chlorinated compounds occurs via a two-step beta elimination pathway through acetylene and chloroacetylene as suggested by Sivavec, et al., (14) in 1996 at a meeting of the RTDF in San Francisco. This not-so-subtle change is important in understanding the possible ramifications of incomplete treatment due to not meeting groundwater residence times within the PRB system. We also now have field evidence that these systems, while generally abiotic within their core (15) also promote significant evolution of dissolved hydrogen gas (16) and likely promote reductive and treatment capacity somewhat outside of therigorouslimits of the PRB itself (17, 18). Similar issues in understanding reaction mechanisms also become important in evaluating the longevity of the PRB. As the number of PRBs that have been operating for 3 to 4 years or more increases, the amount of performance data along with the results of special studies designed to assess the potential for porosity loss due to mineralization and particle blinding within the PRB, allows the development of methods to ensure that the treatment reactions are maintained. Similar to the serendipitous, or rather fortuitous discovery by students and faculty at the University of Waterloo in the mid to late 1980s that reduced met-



43

Figure 3. History of PRB Publication Activity 1991 - 2000.

Note: Data compiledfromthe Oregon Graduate Institute iron references data base. Therefore, numbers indicate publication activity for topics focused on use of zero-valent metals for contaminant treatment



44

als, chiefly iron, can degrade chlorinated hydrocarbon compounds in an aqueous system, the approach to developing treatment media for use in PRBs has been less systematic than pragmatic. That is, the treatment media selection process has relied much on an understanding of basic chemical principles that has evolved over the centuries. For example, the basic processes responsible for iron corrosion in an aqueous system, carbonate equilibria response to pH shifts, sorption mechanisms associated with the amount of available organic matter, and ion exchange reactions, form the basis of treatment media selection and application within a PRB. In most cases, the basic chemistry appears to be understood well enough to allow a PRB design to meet treatment objectives. We know that mineralization and corrosion will occur within an iron-based PRB; field data indicate that most precipitation occurs at the PRB interface for many constituents (e.g., iron oxyhydroxides), and is more uniform over the flow through thickness of the PRB for other constituents (19). The results of research also continue to show progress in designing and selecting new mixtures of treatment media that (a) promote the necessary reactions for remediating a greater variety of constituents in a given plume (i.e., the use of the zeolite clinoptilolite [see Figure 2] for removing certain inorganic constituents such as strontium from groundwater), and (b) are less prone to short-term decreases in reaction rates and fouling. For PRBs designed to promote sequential treatment of a chemical plume (e.g., mitigating the migration of inorganic constituents through geochemical reduction, followed by aerobic treatment of aromatic hydrocarbon constituents), the design must consider methods to reduce or eliminate potential competition between the multiple treatment reactions (e.g, high pH conditions from upgradient zero-valent iron zones may limit the ability of certain downgradient oxygen releasing systems to function as intended). Other design considerations also are required to assure hydraulic efficacy between sequential treatment reactive cells or zones. The use of a PRB combined with an approach that relies on natural conditions to provide further degradation of target chemicals downgradient of the PRB (i.e., natural attenuation) will likely be a much more common remedy as the relevant chemical reactions (with respect to both systems) become better understood. The hydraulic control system is critical to the ultimate performance and effectiveness of the PRB. The hydraulic system functions to (a) route the affected groundwater through the treatment matrix at an appropriate flow rate, and (b) prevent the migration of untreated groundwater around the treatment zone. The hydraulic control system may be simply the ambient groundwater flow system, or it may be augmented by lateral and/or horizontal low permeability barriers constructed to control groundwater pathways and velocity.


45

The hydraulics of a PRB can be affected by a number of conditions that reflect both the ambient nature of the site and the effects of PRB construction on the inherent hydraulic system. These conditions can include: •


• • • •

Effect of regional and local precipitation on hydraulic gradient conditions (direction and magnitude). Surface water infiltration in vicinity of the PRB. Groundwater pumpingfromnearby areas. Smearing or alteration of major groundwater flow zones due to PRB construction. Compaction of the treatment material and subsequant loss of permeability within the PRB.

The presence of one more of these conditions can result in unintended hydraulic performance. It has become apparent that over the past decade of PRB deployment, those PRB systems that have not performed as designed can attribute most of the difficulties to a lack of hydraulic performance rather than to chemical processes. Primary causes for the lack of performance likely are due to incomplete hydraulic characterization of a site, and thus incomplete designs. Most designs to-date would appear to have regular geometries over the entire length of a PRB alignment (e.g., the reactive cell of a long, that is hundreds to thousands of foot-long, PRB system may not be designed to accommodate variability in hydraulic characteristics) which can not effectively respond to changes in ambient groundwater velocities or pressure heads, or a PRB may not have been installed to fully penetrate an underlying low permeability zone along its entire length, thus allowing the potential for underflow to occur. Pilot tests have been one method used to assess the performance of a PRB, and gain valuable design information prior to investing the capital for a full-scale system. However, the pilot test itself can be prone to unintended hydraulic performance, and thus the results can bias (either for or against full-scale installation) the ultimate decision toward the final remedy. If a pilot scale test is applied, it is recommended that the direction and magnitude of the hydraulic gradient be controlled (e.g., potential changes in flow direction are kept to a minimum and the pilot system is oriented to capture the flow as close to perpendicular to the ambient flow direction as possible) to avoid the possibility that groundwater flow would be diverted around the, typically, very short-length of the pilot PRB. Also, the longevity of the pilot test should be sufficiently long to allow a representative evaluation of the treatment process to be made. The pilot-system



46 installation process itself stresses the subsurface system and causes transient changes to hydraulic and geochemical conditions. The duration of the pilot test should be sufficient to allow these initial effects from construction (e.g., transient water level changes and input of atmospheric oxygen) to dissipate. The reaction process also may require "conditioning" within the subsurface system to allow approximately "steady-state" conditions to develop. A review of the results of numerous pilot and bench-scale tests suggests that more than 20 pore volumes of flow through a PRB composed of zero-valent iron may be required before the reaction-rate can be accurately assessed (J. Vogan, 2001, personal communication). This conditioning process, however, may complicate the assessment of early system performancem, however, as an evaluation of both pilot- and full-scale systems also indicates that treatment media can age (i.e., the reaction rate can decrease) as the number of pore volumes increases (A. Gavaskar, 2001, personal communication).

E M P L A C E M E N T AND COSTS The future of PRB use is tied directly to two parameters that each potential application must consider: emplacement approach and cost. Most PRBs to-date have been installed to shallow depths (e.g., less than 50 feet depth) using conventional excavation/fill or trenching techniques. Some of these methods have used sheet piles, trench boxes, or biodegradable slurry solutions to maintain the open trench during infilling of the treatment matrix. Deeper installations (to approximately 120 feet deep) have attempted deep soil mixing, jetting, injection, and fracturing (pneumatic and hydraulic) to emplace the treatment matrix. These developing methods also could be used to place treatment material in depth-discrete chemical migration pathways. Certainly, maintaining the hydraulic control and the groundwater residence time requirements in the deeper installations will be more complicated in the more shallow installations and will require more comprehensive site characterization and design studies to assure a successful application. However, many of the deeper groundwater plumes have few remedies available other that conventional pump-and-treat. Consequently, the development of the more advanced emplacement approaches during the corning years should make the PRB concept more widely acceptable for the plumes in the deeper and more complex aquifer settings. Cost data on most PRB installations to-date are available only in a general fashion, although general unit costs for the various components of a PRB can be accurately assessed (for example, the average cost of zero-valent granular iron is approximately US$350 to $400 per ton, while hydraulic slurry systems for hydraulic control may have a cost of approximately US$10 per foot). However, the



47 overall costs, which include characterization, design, legal, permitting, and other issues are not well characterized. Generally, the capital cost of a PRB should be in the same range as the capital for many other active groundwater remedies, including groundwater extraction with above-ground treatment ("pump-andtreat"). The savings that a PRB application affords primarily is through lower costs of operation and maintainance activities, compared to the conventional alternatives. As presented during the recently completed U.S. Envionmental Protection Agency training session on PRBs (7), even with then need to replace a PRB on a 10-year interval, the long-term costs for the PRB approach should remain considerably less than the pump-and-treat alternative. As an example, installation of the Sunnyvale PRB (15), which remains the longest running commercial system, reduced annual project expenditures from more than approximately $300,000 to approximately $50,000 following replacement of the former pump-and-treat system at the site. The investment of the capital expenditure for this early PRB, which was on the order of approximately $lmillion, was thus returned in project cost savings after only a few years of operation (as a note, in 1994 when this PRB was constructed, the unit cost of zero-valent iron was approximately $750 per ton, nearly twice today's price). The annual costs at the site are expected to continue to decrease in the future as well.

LESSONS LEARNED AND FUTURE CONSIDERATIONS The success of a PRB in meeting the project specific treatment objectives relies on the successful implementation of both the chemical treatment and hydraulic control systems. Significant work continues in the areas of developing enhanced chemical treatment media for a wider variety of target contaminants and under a wider variety of ambient hydrogeological and hydrochemical conditions, and for increasing the longevity of these materials. We anticipate that during the next ten years, research and development of remedial technologies will progress such that most currently known environmental contaminants will be treatable by some reactive media. We also would anticipate that our understanding of the longevity of these treatment measures within a PRB system will also have evolved considerably (20). Based on a review of case studies, conference proceedings, and technical documents, most of the emphasis over the last 10 years with regard to PRB development, implementation, and monitoring, has been on the chemical treatment portion of the PRB system, with less discussion on how best to design the PRB to incorporate the PRB hydraulics with the chemical processes. Design considerations that more thoroughly address the hydraulic conditions will assure American Chemical Satiety Library

1155 îetfisu&w.

Washington, D £ 20MB Henry and Warner; Chlorinated Solvent and DNAPL Remediation ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

48 that the chemical treatment processes are as effective as possible in treating affected groundwater.


During the past ten years, the concept of the PRBs as a groundwater remediation alternative has developed from being a hypothetical to a practical solution for many sites. The PRB has been demonstrated to be cost effective for specific sites. We now have more than 5 years of performance data from several sites by which we can better design future PRB installations. From these and more recent installations it is recognized that key aspects to assuring the success of a PRB application include: • • • • • •

Developing a comprehensive understanding of site characteristics including chemical distribution and hydrogeology. Completing a more comprehensive assessment of the relationship between hydraulic design and efficacy of the treatment process. Designing the PRB system to optimize longevity. Developing monitoring programs that provide representative and accurate performance information. Continuing to educate potential users and regulators on the key design aspects of PRB systems. Developing a comprehensive cost database for the variety of PRB systems deployed.

Following the April 2001 ACS special symposium on PRBs, a special workshop was convened on April 5, 2001 to discuss the future of PRB technology and to list those considerations that will be paramount to assuring that PRBs are designed effectively and can address sites with greater complexity. The workshop identified the following topics that would appear to be of most interest to the industry at large: Treatment approaches -sequential treatment, alternative media Emplacement methods - deeper and less expensive methods Longevity and methods to rehabilitate PRBs Assessing causes for unintended performance of PRBs Performance monitoring techniques and strategies Developing PRBs for the future Within the general discussion of the above topics, it was generally agreed to by the participants that the PRB concept remains a valid and potentially powerful tool for effectively mitigating contaminant plumes at a wide variety of sites


49


and protecting potentially threatened downgradient users and resources. However, the group agreed that more work is required to both (a) understand the design criteria necessary to implement a successful remedy, and (b) promote the potential use of the PRB concept at complex sites. As a final note, the group also agreed that collection and interpreting performance information on those PRB sites with unintended results are as important as collecting information on the successful applications so that effective guidance on PRB design and application can be properly formulated and circulated throughout the groundwater remediation industry.

REFERENCES 1. 2.

3. 4. 5.

6.

7. 8. 9. 10.

11. 12.

McMurty, D.C.; Elton, R.O. Environmental Progress. 1985, Vol. 4, No. 3, 168-170. Gillham, R.W.; Burris D.R., Subsurface Restoration Conference, 3 International Conference on Ground Water Quality Research, Dallas, 1992, 66-72. Gillham, R.W.; O'Hannesin, S.F. Ground Water, 1994, Vol., 32, No. 6, 958-967. O'Hannesin, S.F.; Gillham, R.W. Ground Water, 1998, Vol. 36, No. 1, 164170. Yamane, C.L.; Warner, S.D.; Gallinatti, J.D.; Szerdy, F.S.; Delfino, T.A.; Hankins, D.A.; Vogan, J.L. 209th National Meeting, American Chemical Society, Anaheim, California, 1995, Vol. 35, No. 1, 792-795. Szerdy, F.S.; Gallinatti, J.D.; Warner, S.D.; Yamane, C.L.; Hankins, D.A.; Vogan, J.L. ASCE National Convention, American Society for Civil Engineers, Washington, D.C., 1996, 245-256. U.S. Environmental Protection Agency, Treatment Technologies for Site Cleanup, Annual Status Report (Ninth Edition), 1999. Warner, S.D.; Szerdy, F.S.; Yamane, C.L. Contaminated Soils, 1998, 3, 315-327. Gee, G.W.; Wing, N . R., Thirty-Third Hanford Symposium on Health and the Environment, 1994, Vols. 1 and 2. Taylor, T.P.; Sauer, N.N.; Conca, J.L.; Streitelmeier, B.A.; Kaszuba, J.P.; Jones, M.W.; Ware, S.D. Groundwater Quality 2001, Third International Conference on Groundwater Quality, University of Sheffield, United Kingdom, 2001, 110-112. U.S. Environmental Protection Agency, EPA542/B-00/001, Office of Solid Waste and Emergency Response, Washington, D.C., 2000. U.S. Department of Energy, Grand Junction, Research and Application of Permeable Reactive Barriers. April 1998. rd



50

13. U.S. Environmental Protection Agency, EPA/600-R-98/125, Office of Solid Waste and Emergency Response, Washington, D.C., 1998. 14. Sivavec, T.M.; Mackenzie, P.D.; Horney, D.P.; Baghel, S.S. International Containment Technology Conference, St. Petersburg, Florida, 1997, 50. 15. Warner, S.D.; Yamane, C.L.; Bice, N.T.; Szerdy, F.S.; Vogan, J.L.; Major, D.W.; Hankins, D.A. Proceedings of the First International Conference on Remediation of Chlorinated and Recalcitrant Compounds, Monterey, California, 1998, 6, 145-150. 16. Sorel, D.; Warner, S.D.; Longino, B.L.; Hamilton, L.A. Preprints of Papers, American Chemical Society, Division of Environmental Chemistry, 2001. 17. Puls, R.W.; Paul, C.J.; Powell, R.M. Applied Geochemistry. 1999, Vol. 14, No. 8, 989-1000. 18. Weathers, L.J.; Parkin, G.F.; Alvarez, P.J. Environ. Sci. Technol., 1997, 31, 880-885. 19. Phillips, D.H.; Gu, B.; Watson, D.B.; Roh, Y.; Liang, L.; Yee, S.Y. Environ. Sci. Technol., 2000, 34, 4169-4176. 20. Battelle, Monitoring plan for the evaluation of permeable reactive barriers at multiple Department of Defense sites. Prepared for NFESC, Port Hueneme, California, 2000.


Ten Years of Permeable Reactive Barriers ... - ACS Publications

Recommend Documents